Skip to content

[quality] Shell scripts and Python hooks have zero unit test coverage #553

Description

@kubestellar-hive

Quality Improvement Epic — projectbluefin/common

This epic tracks the systematic improvement of test coverage and static analysis across all shell scripts and Python hooks shipped in the Bluefin common layer.

Background

An architectural review (2026-06-10) identified that the repo grew a test harness around libsetup.sh incrementally but never established a testing contract for new code. The result: security-critical and user-facing scripts accumulated with zero test coverage. This epic closed that gap.

Full analysis: docs/quality/ARCHITECTURE-REVIEW.md


Work Completed

PR What shipped Status
#540 20 pytest tests for hooks.py + 18 bats tests for libsetup.sh + setup scripts + CI workflow ✅ Merged
#554 Shellcheck expanded from 1 → 11 scripts + 3 profile.d fixes ✅ Merged
#568 27 new bats tests (bling / privileged-setup / luks-tpm2) + docs/TESTING.md + 80% coverage gate ✅ Merged
#573 bats tests for rechunker-group-fix, ublue-bling-fastfetch, changelog.just + --cov-fail-under=80 🟡 Awaiting review
#579 bats tests for ublue-fastfetch (8 tests) + CI wiring 🟡 Awaiting review
#580 Fix changelog bats CI failure (IMAGE_INFO_FILE override) 🟡 Awaiting review (fixes #573 CI)
docs push (2026-06-10) TESTING.md: remove stale open.sh exemption, add ublue-motd + image-info exemptions, fix coverage target ✅ On main

Acceptance Criteria

Child Issues (closed)

Metadata

Metadata

Assignees

No one assigned

    Labels

    1-triageNew work awaiting human triage.

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions